IGMPv3 hot interface detach panics?

[Sam Leffler]

Bruce Simpson wrote:
> Thomas Sparrevohn wrote:
>> ...
>> I am getting a panic as well - but I am not sure its related
>>
>>   
>
> Looks like inm is NULL at line 1706 in igmp.c
>
> Can you confirm this by inspecting the core dump? i.e.
> frame 9
> print inm
>
> gcc has inlined the functions at this point. It may be a symptom of a 
> race.
> Can you reproduce what the system was doing at the time?

It's the same panic I hit and not a race.  I caught it just 
single-stepping through a normal net80211 vap teardown on device eject.  
Not sure I'll have time to look at it today.

    Sam